1. Field of the Invention
The present invention relates generally to the field of the humoral immune defense against pathogens that enter the body through mucosal surfaces. More particularly, it concerns the field of recombinantly produced dimeric IgA antibodies and the secretion of such antibodies into the mucosal secretions for protective immunity from external pathogens including viral, bacterial, bacterial toxins and macroscopic parasites.
2. Description of Related Art
The mucosal surfaces of the body represent the largest area of exposure of the body to external pathogens, 400 m.sup.2 compared to only 1.8 m.sup.2 of skin area (Childers et al., 1989). IgA is the immunoglobulin subclass primarily responsible for humoral immune protection at this large exposed surface. IgA can intracellularly associate with J-chain via a cysteine in the C terminal "tail" to form dimeric IgA (dIgA, Koshland, 1985) which can be bound by the polymeric immunoglobulin receptor (pIgR) and transported across mucosal epithelia (Mostov and Blobel, 1982; Underdown, 1990; Mostov, 1994) to be released as secretory IgA (sIgA). Therefore sIgA serves as the first line of humoral immune defense at mucosal surfaces (Underdown and Schiff, 1986).
IgA is a tetrameric protein comprising two identical light chains (.kappa. or .lambda.) and two identical heavy chains (.alpha.) which endow IgA with its biologically specific properties. In the human there are two IgA isotypes, IgA1 and IgA2, present as a result of the duplication and subsequent diversification of a large segment of the human heavy chain locus (Torano and Putnam, 1978, Putnam et al., 1979, Tsuzukida et al., 1979,). The overall domain structure of IgA appears to resemble IgG in that it contains three constant domains (C.alpha.1-C.alpha.3), with a hinge region between the C.alpha.1 and C.alpha.2 domains. All IgA isotypes (as well as IgMs) have an 18 amino acid "tailpiece" C-terminal to the CH3 domain not present on IgG which enables polymeric Ig formation (Garcia-Pardo et al., 1981, Davis et al., 1988).
Serum IgA is present in monomeric or polymeric forms, which are mostly dimers, though tetramers and higher polymers have been observed. Polymerization is unique to antibodies of the IgA and IgM isotypes and is mediated by the tailpiece in conjunction with the J chain, one molecule of which is present per dimer of IgA (Zikan et al., 1986). The J chain is a small glycoprotein (15 kD) component of polymeric IgA and IgM (Koshland, 1985). J chain contains six interchain disulfide bonds and two additional cysteine residues at positions 15 and 79 which form disulfide bonds to the penultimate cysteine of one heavy chain in each IgA monomer subunit (Bastian et al., 1992). Thus J chain bridges the two IgA monomers, while the other two tailpiece cysteine residues bind each other directly, stabilizing the dimeric molecule further.
Dimeric IgA (as well as polymeric IgM) is specifically bound by the pIgR, expressed on the basolateral surface of mucosal epithelial cells, and transported through these cells to be secreted at the mucosal surface. Secretory IgA (sIgA) retains most of the extracellular region of the pIgR, termed secretory component (SC), covalently bound to one of the IgA monomers (Underdown et al., 1977).
Studies in J chain knockout mice show that J chain is not absolutely necessary for IgA polymerization although it is much less efficient in the absence of J chain (Hendrickson, et al., 1995). In these mice levels of IgA in bile and feces were much reduced, as was the level of dimeric IgA in serum. Further investigation revealed that intestinal, mammary and respiratory secretions contained IgA in a predominantly monomeric form (Hendrickson et al., 1996), therefore J chain is not necessary for IgA secretion but appears to stabilize the interaction between secreted IgA and secretory component.
The structures on the pIgR which mediate the interaction with IgA have been partially identified as has the mechanism of association between IgA and pIgR. The pIgR is a 110 kD transmembrane glycoprotein with five immunoglobulin superfamily homology domains (I-V) in the extracellular region (Mostov et (et al., 1984, Eiffert et al., 1984). The primary site of interaction of pIgR with dIgA is in domain I (Frutiger et al., 1986), which participates in a high affinity (10.sup.8 M.sup.-1), non-covalent interaction (Kuhn and Kraehenbuhl, 1979). Further mapping of the dimeric IgA binding site within domain I of the pIgR has identified a peptide comprising residues 15-37 of human pIgR which binds dIgA (Bakos et al., 1991a). A mutational approach, based on modeling of the domain I sequence on known Ig variable domain structures, demonstrated that the loops in analogous positions to the three V region CDRs made up the dIgA binding site (Coyne et al., 1994). This suggests that the interaction of the pIgR with dIgA is similar to the interaction of antibody with antigen, or to be more precise, the interaction of a single V domain with antigen.
The second stage of the interaction between pIgR and dIgA involves covalent binding of domain V to the Fc of one of the subunits in dIgA (Lindh and Bjork, 1974, Cunningham-Rundles and Lamm, 1975). This single disulfide is formed between cys467 in domain V of secretory component and cys311 located in the C.alpha.2 domain of a heavy chain in one IgA subunit (Fallgreen-Gebauer et al., 1993). Disulfide formation appears to be a late event in the secretion pathway and is not absolutely necessary for transcytosis (Chintalacharuvu et al., 1994, Tamer et al., 1995).
Protective antibodies of the IgA isotype have been documented against a wide range of human pathogens including viruses such as IIIV (Burnett et al., 1994) and influenza A (Liew et al., 1984), bacteria (Tarkowski et al., 1990, Hajishengallis et al., 1992) bacterial toxins and macroscopic parasites (Grzych et al., 1993). There are several mechanisms by which IgA exerts its antimicrobial effect and they may be divided into active (e.g. Fc receptor binding or complement activation) and passive (e.g. blocking of viral receptors for host cells or inhibition of bacterial motion) mechanisms.
A number of studies have demonstrated the association between strong mucosal IgA responses and protection against viral infection with rotavirus (Underdown and Schiff, 1986, Feng et al., 1994), influenza virus (Taylor and Dimmock, 1985, Liew et al., 1994), poliovirus (Ogra and Karzon, 1970), respiratory syncytial virus (Kaul et al., 1981), cytomegalovirus (Tamura et al., 1980) and Epstein-Barr virus (Yao et al., 1991). Secretory IgA is therefore, successful in preventing these viruses from gaining access to the body by blocking infection at the site of entry, namely the mucosal surface. Passive immunotherapy with intranasal IgG Fabs was protective against respiratory syncytial virus (Crowe et al., 1994), showing that the mere presence of neutralizing anti-viral antibodies, without any effector function, at the mucosal surface can prevent viral infection.
Due to its complex interactions with the host immune system the HIV virus has proved very difficult to contain once it has entered the body. Clearly a strategy based on exclusion of the virus from the body would be ideal. Mucosal IgA antibodies offer this possibility and have a key role in many viral infections. Since the mucosal surfaces of the body are the usual point of entry of the virus to the body it seems logical to concentrate some effort at developing this first line of anti-viral defense. There are several reports suggesting a protective role for passive immunization in patients with AIDS.
It has been observed that HIV-1 infected individuals had neutralizing antibodies and high titer anti-viral antibodies in contrast to AIDS patients, who had low levels of anti-viral antibodies (Karpas et al., 1988). Passive immunization of both ARC and AIDS patients had beneficial effects. Neutralizing anti-HIV antibodies have been produced using combinatorial libraries derived from long term asymptomatic HIV infected donors (Barbas et al., 1992 and 1993), suggesting a role for humoral immunity in limiting progression of HIV infection. Passive immunotherapy, using human anti-HIV sera, has been shown to delay the progression of disease in HIV-infected patients (Vittecoq et al., 1995). Vaccination studies in macaques have suggested that secretory IgA can play a major role in neutralizing HIV-1 (Bukawa et al., 1995). The relative roles of systemic and mucosal IgA in the anti-HIV response remains to be elucidated though neutralizing IgA antibodies are present in the serum of HIV infected patients (Burnett et al., 1994). The presence of sIgA antibodies in the saliva of HIV-1-infected individuals correlated well with asymptomatic HIV infection, whereas the patients with AIDS displayed reduced sIgA levels in saliva (Matsuda et al., 1994). In contrast, no such correlation was observed with the serum IgG in these individuals.
The presence of maternal serum IgA against HIV has proven to be of prognostic value in determining the materno-fetal transmission (Re et al., 1992). In this study mothers who gave birth to uninfected children had serum IgA against HIV, directed particularly against the gp24 protein. However the mothers of infected children did not possess this reactivity. It has been suggested that some materno-fetal transmission of the HIV virus occurs during the process of parturition (Livingston et al., 1995). Therefore the presence of this serum reactivity suggests that secretory IgA (sIgA) with this specificity may confer protection during the process of birth by neutralizing the HIV present in the birth canal. The type of IgA response to the virus seems to be specific in that IgA1 is preferentially expressed (Kozlowski et al., 1992) and perhaps this subclass of IgA, is more effective at combating the virus. These studies demonstrate the potential for antibodies in combating the disease and give rise to the hope that vaccination, in non-immunocompromised individuals, may elicit protective antibody-mediated immunity. Unfortunately, there is no immunotherapy available to treat the HIV at the mucosal surface, before it enters the body.